Typography and Grammar in Technical Paper Writing
A Start-Up Guide for Research Students

### Units

1. Accuracy in the prefices of units are often overlooked. Specifically, the following prefices are often set in the wrong case.
• The "k" in kilo should be in lower case, e.g., km, kg, kHz, kΩ.
• The "M" in 'Mega' should be in upper case, e.g., MHz, MΩ.
• The "m" in 'milli' should be in lower case, e.g., mW, mH, mF.
• In computer terminology, 'kilobyte' is commonly abbreviated as KB, with "K" in upper case. This "K", however, means 1024!
2. When units are typed in complete words, always use lower-case letters.
• The unit of power is "watt" or "W", not "Watt".
• The unit of time is "second" or "s", not "sec" or "Sec". (NOTE: "S" is the unit of conductance.)
• The unit of resistance is "ohm" or "Ω", not "Ohm", etc.

### Typefaces for variables and units

1. Variables and math symbols should be set in italics. For example, (A + 5B) + 8C = 0. Note that when typesetting equations, numerals, brackets, signs and operators should be set in roman.
2. Units should be set in roman typeface. For example, 10 kHz, not 10 kHz!
3. Words in subscripts should be set in roman typeface if they have a linguistic meaning. For example, Vref and Iout are preferred over Vref and Iout . In LaTeX, type $V_{\rm ref}$ to generate Vref.

### Grammar

1. Countable nouns do not appear in singular form without an article ('a' or 'the') in front. Use plural form if it has a collective meaning. The following is incorrect:

X  High-frequency amplifier has cascode stage that avoids Miller bandlimiting.

Here, two articles are missing. A "The" (definite) or "a" (indefinite) should be used in front of "High-frequency amplifier" and "cascode stage". In this example, if "High-frequency amplifier" appears for the first time in the text, "A" should be used. But if it has already been explained or defined, "the" should be used. Moreover, for "cascode stage", the choice is obviously "a". Alternatively we may have the following grammatically correct version (but not technically correct though):

High-frequency amplifiers have cascode stages that avoid Miller bandlimiting.

2. Check multiple occurrence of verbs within a sentence. For example, the following is incorrect.

X  Assume that the input resistance is infinitely large, we can find the gain of the amplifier readily.

Here, 'Assuming' should be used instead of 'Assume'. Alternatively, you can say "Assume that the input resistance is infinitely large. We can readily find the gain of the amplifier."

3. Be precise about pronouns. 'It', 'they', 'these', etc. should be used without confusion.
4. Be precise in using adverbs and adjectives. For example, the following is incorrect (seen many times in bad papers!):

X  The circuit performs its function satisfactory. --- (satisfactorily)
X  This amplifier does not necessary require a high-impedance input. --- (necessarily)

### Logic and consistency

1. Symbols used should be unified throughout the paper, and should be consistent with those shown in figures.
2. Keep the structure consistent for all items in a list. For example, the following is incorrect.

X The aim of the design is

• to provide a non-inverting gain of 20 dB;
• the high-frequency roll-off should be at least 5 MHz.

Obviously, the second item in the bullet-point list above does not match the phrase "The aim of the design is", and should be corrected as "to achieve a high-frequency roll-off of at least 5 MHz."

3. Any abbreviation should be explained when it appears for the first time. For example,

The power factor (p.f.) of the circuit is equal to 0.7. If a capacitor is connected in parallel to it, the p.f. increases to 0.9.

4. Multiple-word nouns should be hyphenated when used as adjectives.

This circuit achieves power factor correction.
This is a power-factor-correction circuit.

5. Texts and symbols that appear in figures should be similar in size to those appear in the main text body.

6. Pay attention to the language logic. In the following example, "parasitic capacitance" cannot be a cause of anything, but its presence is.

X  The high level of emitted noise is due to parasitic capacitance. --- (the presence of parasitic capacitance)

### Using "respectively" appropriately

• The word "respectively" must always connect several pairs of items. In general, the structure is

X, Y, and Z are connected to A, B, and C, respectively.

The word "respectively" cannot be used to refer two or more things without the logical pairing. A common mistake is like the following (often translated from Chinese):

X  The two systems are analyzed in terms of their small-signal and large-signal behavior respectively.

In the above sentence, "respectively" is inappropriately used. In fact, the word can be removed if the sentence actually means that both systems are to be analyzed in terms of both small-signal and large-signal behavior. However, it could have meant "System A and system B are analyzed in terms of their small-signal and large-signal behavior, respectively." Here, clearly, system A is analyzed in terms of its small-signal behavior, whereas system B is analyzed in terms of its large-signal behavior. The pairing should be explicit and unambiguous.

### Consistent formatting of bibliography

1. The problem is carelessness! Check the formatting requirement of the journal or conference proceedings concerned and make sure all references are formatted consistently and in a unified way.
2. Look how ugly the following is!
• References:
1. C.K. Chan and D. C. Tan, "Paper title is Here," IEEE Transactions on Power electronics, Vol. 7, number 10, pp. 1-10, Oct. 2012.
2. Chan C.K. and David C. Tan, "Another Paper Is Here," IEEE Trans. power Electron., 7(10): 8-20, 2013.
3. Chan, Charles K, The book title is here, Wiley, New York, 2009.
4. D.C. Tan, The Book Title, Englewood Cliffs: Prentice Hall, 2001.
Here, "D. C. Tan" and "D.C. Tan" have different formats. The same journal name appears differently in two references. "Vol." and "number" are set inconsistently. The volume and issue numbers in references 1 and 2 are also inconsistent. Two different book title formats are used in references 3 and 4. And you can find at least 5 more problems if you are observant enough!
The problem has nothing to do with typo or grammar. It is pure personal management.
A good researcher begins with careful writing and meticulous attention to details.

More examples of wrong logic and grammar can be found in http://www.eie.polyu.edu.hk/~cktse/TechnicalWriting.pdf which was presented to our DEng students as part of the curriculum.

Written by Michael Tse, 14 November 2001; revised 20 July 2012.